LBM simulation for CO2 saturation monitoring from elastic velocity and resistivity: Migration of supercritical CO2 in porous media under several PT conditions

In Carbon dioxide Capture and Storage (CCS), the monitoring of injected CO2 is critical for (1) predicting the risk of CO2 leakage from storage reservoirs, (2) reducing the cost and increasing the efficiency of CO2 injection, (3) reducing the risk of injection-induced seismicity, and (4) reducing the risk of creating new fractures, and hence opening new flow paths in an otherwise low-permeability capping formation. In seismic approaches, the change in seismic velocity derived from time-lapse seismic surveys can be used to evaluate the distribution of injected CO2, because the P-wave velocity decreases dramatically as the CO2 starts to invade the pore space of a rock initially saturated with brine. However, the rate of change decreases considerably when the saturation of CO2 reaches a value of 20% making the estimation of higher saturations a difficult task. Usefulness of resistivity to estimate CO2 saturation in wide porosity range was presented from laboratory experiments; the resistivity value increases with CO2 saturation even in higher saturation range. However, we cannot characterize the CO2 distribution within pore space only from laboratory experiments. The detailed investigation is needed to generate quantitative description of CO2 migration in porous media and to construct relationship between CO2 saturation and field-derived properties (i.e., seismic velocity and electric resistivity) for the quantitative monitoring of injected CO2. In this study, we apply two-phase lattice Boltzmann method (LBM) to the digital rock models, in order to investigate (a) seismic velocity and (b) electric resistivity under several pressure-temperature (PT) conditions. The simulation study has been compared with laboratory experiments. LBM is one of the computational fluid dynamics methods. In this algorithm, fluid as well as supercritical CO2 is represented as aggregation of imaginary fluid particles, then the movement of microscopic and discretized particles is calculated. After we simulate supercritical CO2 injection into water-saturated porous media using LBM, we calculate (a) seismic velocity and (b) resistivity from the estimated supercritical CO2 distribution within pore spaces. From these calculations, we obtain seismic velocity and resistivity for various pore structures at several PT conditions. Our preliminary results demonstrate that electric resistivity of the well-sorted grain model is affected by PT conditions. Because the viscosity and density of supercritical CO2 is much changed by PT conditions, CO2 distribution within pore space is different for each PT condition and influences to the resistivity value. Although the estimated relationship between CO2 saturation and resistivity can be modeled by Archie's equation, the parameters used in the equation should be changed by considering PT condition of the injected reservoir. Therefore, the realistic CO2 distribution within pore space should be considered for the quantitative geophysical monitoring. In this study, we further calculate relative permeability for each condition.